On the Suitability of Reinforcement Fine-Tuning to Visual Tasks
This study addresses the applicability of RFT to visual tasks for researchers in multimodal learning, providing early insights but is incremental as it builds on existing RFT methods.
The paper investigates the suitability of Reinforcement Fine-Tuning (RFT) for visual tasks, finding that RFT generally outperforms Supervised Fine-Tuning (SFT) on these tasks, with benefits from increased reasoning on complex tasks but drawbacks on simple ones.
Reinforcement Fine-Tuning (RFT) is proved to be greatly valuable for enhancing the reasoning ability of LLMs. Researchers have been starting to apply RFT to MLLMs, hoping it will also enhance the capabilities of visual understanding. However, these works are at a very early stage and have not examined how suitable RFT actually is for visual tasks. In this work, we endeavor to understand the suitabilities and limitations of RFT for visual tasks, through experimental analysis and observations. We start by quantitative comparisons on various tasks, which shows RFT is generally better than SFT on visual tasks. %especially when the number of training samples are limited. To check whether such advantages are brought up by the reasoning process, we design a new reward that encourages the model to ``think'' more, whose results show more thinking can be beneficial for complicated tasks but harmful for simple tasks. We hope this study can provide more insight for the rapid advancements on this topic.